PPL controll of a LIPD FM TXer.


Senior Member
Ive been experimenting with a low power FM band TXer. See here http://electronics-diy.com/tx200.php.

I have built crystal controlled FM txer's , they wander on the band with temp changes. Apparently PPL's stay rock solid on the TX freq they are set at.

The TXer can be digitally tuned by using a digital to analogue converter under digital controll. As far as I can understand it the mcu takes a reading from the Txed signal and then adjusts the digital to analogue converter output voltage if need be to put the TXer back on the freq it should be on.

The design used in this example uses a PIC16F84 and a SAA1057. The SAA1057 data sheet is here.

I got to thinking it might be possible to do this sort of thing with a picaxe and a digital to analogue converter IC.

The biggest problem I can see straight off is can the picaxe count frequencies in the FM band range 88- 107 mghz. I dont see how this could be done because the picaxe isn't running at any where near that speed.

Can this be done? Any input would be appreciated.

People often use PWM instead of DACs so perhaps that might be of use.

Be aware that the picaxes operate at low clock speed so I should think there is no way they could count pulses in the FM band.

What about using some sort of demodulation method and then use the picaxe at that stage?


Technical Support
Staff member
The principle is to measure the frequency, check if it's too fast or to slow and adjust as necessary.

It shouldn't matter that the PICAXE can only read slowish signals, because the same percentage error is in a divided down frequency as in the original. In fact there are benefits to this because that helps to reduce inaccuracies ( resolution ) when reading the error.

If you had a motorised clock driven from an RC oscillator, you could check it every 60 minutes then tweak the R potentiometer up or down to get it running at the right speed. It's much easier to see/measure a 1 second error to a good degree of accuarcy after an hour than it is a 300nS error every mS.

The challenge is in measuring and processing with enough accuracy that reading that any adjustment to the oscillator is subtle and doesn't cause it to jump wildly, while adjusting frequently enough to keep the error getting too great. You'll also need a good, stable analogue signal or the oscillator will drift as the analogue ripples.


Senior Member
If you're going to add a PLL frequency control system to your transmitter, you probably won't have to worry about stability too much. Your (relatively) low frequency reference oscillator crystal determines the frequency stability of your system. A crystal oscillator operating at something like 10 MHz should be very well behaved, and your FM band carrier frequency will inherit the crystal oscillator's stability.

If you want to use a Picaxe with your PLL system, it would be ideal to convert a human-friendly frequency tuning system into the divide-by-n codes used by PLL chips.

Although it probably doesn't have a home in your system, search for "Huff n' Puff" for a very simple, but really weird, way of step-locking an oscillator. While I don't think a Picaxe is really the best vehicle for a huff n' puff system, you might gain some ideas which might help you engineer your transmitter.

Have fun! Sounds like an interesting project. PLLs are lots of fun for experimenters.



Senior Member
BrendanP: You perhaps know Andrew "BrightSpark" & I have been considering a simple Picaxe data system using the normal FM band. Compared with good old 433MHZ the appeal of 88-108MHz relates of course to todays FM receiver super abundance for quick tuning to "hear" the data- perhaps as a tone or simple code.

Aside from the chore of even finding your weak transmission (most regions now have FM bands super crowded with powerful music stations),you'll almost always have wandering signals with simple transmitters due to changing temperatures,sagging battery voltages & stray capacitances. Even the TYPES of caps. used can be critical of course. OK- these hassles can be lived with for skinflint setups, but you'll need to go upmarket circuit wise for more demanding applications. Hence suggest you first detail your needs/range/budget etc. THOUGHT: For decent data stability consider the lowly AM band around 1 or 1.6MHz instead ?

EXTRA: Consider hacking universal iPod/MP3 style transmitters such as the Belkin Tunecast (~US$20)? In my experiences these are often static prone,with poor in-car range, but could be OK for fixed sites. Part of the problem relates to US FCC power limits of just a few mW- not an issue in Oz. I hear they even find use for talking real estate promos, so that when parked outside a place you can tune to their sign's displayed FM freq, &amp; hear more details. Talk about clever marketing &amp; PROFITABLE technology! As recently as 2 years back the commercial AM band &quot;Talking House&quot; <A href='http://www.treasurecoastrealtor.com/Talking%20House.htm ' Target=_Blank>External Web Link</a>,or FM Radiosign <A href='http://www.radiosign.com ' Target=_Blank>External Web Link</a> were needed (~US$300 each), but you can now DIY thanks to these FM units. Stan

Edited by - manuka on 21/07/2007 02:13:06


Senior Member
Thanks for everyones input. Im always impressed at the time and effort that people will expend on answering questions on this forum.

Im looking at the ROHM BH1417 IC. Its a FM band stereo TX IC with PLL.

Stan be aware that max FM band TX power in Australia (and I suspect NZ) is 10 MICROwatts. Thats micro not milli. I saw it called milli in a sil chip article. Ive got the LIPD regs in front of me and its micro. Some devices are restricted to pico watts! What can you do with that output power??!!


Edited by - BrendanP on 21/07/2007 06:14:22


Senior Member
Gasp- I'll check myself! MicroWatt power levels are akin to a superhet. receiver local oscillator of course. Hence 10 &#181;W is certainly hard to credit,especially with all the MP3/iPod transmitters plus wireless &amp; kids spy FM mikes around- some of which are obviously quite powerful. Even 433.92 ISM can be 25mW!

Have just checked the regs <A href='http://www.rsm.govt.nz/cms/policy-and-planning/current-projects/radiocommunications/review-of-spectrum-allocations-for-short-range-devices/an-engineering-discussion-paper-on-spectrum-allocations-for-short-range-devices/7-very-low-power-audio-senders ' Target=_Blank>External Web Link</a> myself- may be loopholes with a &quot;tunnel transmitter&quot;!? Here in NZ there are legal licence free LPFM ( Low Power FM) spots at 88.1 to 88.7 and 106.7 to 107.7MHz allowing up to 500mW. These find good use for ethnic,event,campus, &amp; &quot;arrival&quot; radio at ferry terminals etc =&gt; <A href='http://www.low-power-fm-radio.spacejunk.co.nz/lpfm-radio-station-list.asp' Target=_Blank>External Web Link</a>

I'm not up with present global LPFM regs. but have had past experiences with Tx &quot;campus kits&quot; sold by the likes of Veronica (UK) &amp; Ramsey (US) <A href='http://www.ramseyelectronics.com/hk/default.asp?page=amfm ' Target=_Blank>External Web Link</a>. THOUGHT: Suggest you grab an iPod/MP3 style Tx unit from DSE/Jaycar for trials at least, OR even consider AM band work. Stan


Senior Member
As you point out there are many FM txing devices for sale and in common use that exceed 10 microwatts. For low key personal projects I dont think the gestapo is going to role up at your door if you exceed that limit,(as long as the next door neighbour doesnt complain about all the DTMF tones he keeps hearing along with his ABC FM classical music), you might make waves though if you did a sil chip article promoting the concept though!

&quot;The Australian LIPD Class Licence permits up to 10&#181;W transmitter power for wireless audio and auditory assistance (88 to 108MHz) and also for tunnel communications (87.5 to 108MHz).&quot;


Edited by - BrendanP on 21/07/2007 23:55:45


Senior Member
Reading back to your original post I have two comments.
(1) A PLL is only as acurate and stable as the time base being used as the reference freq. If your crystal is drifting badly I would be checking for the correct shunt capacitance and correct parallel resistance. Excessive loading of a crystal will reduce the effective 'Q' of the crystal. This is not desirable!

(2) Phase Locked Loops don't usualy lock to the time base freq. The TB is usually divided down with common decade counters and then fed to the PLL.